CN116999080A - Method and system for creating quantitative positron emission tomography images - Google Patents

Abstract

The invention relates to a method and a corresponding system (1) for creating quantitative PET images of a subject or object, said method comprising the steps of: recording correction information data by means of an imaging device (2); sending the correction information data to a second An evaluation unit generates an attenuation map based on the correction information data, and forwards the attenuation map as inspection data to the second evaluation unit and/or forwards the correction information data as inspection data to the second evaluation unit; by the second evaluation unit based on a plurality of The examination data are evaluated with reference to the data and possible artifacts in the examination data are detected; corrective measures are initialized via the second evaluation unit and/or corrective measures are proposed to the user; corrected attenuation maps are produced based on the corrective measures.

Description

Methods and systems for creating quantitative positron emission tomography images

Technical field

The present invention relates to a method for creating quantitative positron emission tomography images, a positron emission tomography system, a computer program and a non-volatile computer readable medium.

Background technique

Positron emission tomography (PET) is often combined with other imaging methods, especially magnetic resonance tomography (MRT) and computed tomography (CT). PET here offers the possibility of quantifying biologically or medically relevant parameters, for example by administering radiopharmaceuticals and tracking their (possibly time-dependent) quantity and distribution in the body (tracer principle or tracer kinetics). A prerequisite for this is that PET provides images that are quantitative, reproducible and as artifact-free as possible. To produce quantitative PET images, correction factors are usually applied when reconstructing the images. Corrections may include, inter alia, normalization, attenuation correction or absorption and scattering corrections. With the attenuation correction, the attenuation of the signal due to the absorption of photons in the tissue layer is taken into account. The attenuation may differ individually for different patients depending on the respective properties, in particular the material and/or thickness of the tissue. Attenuation correction is often necessary in order to obtain quantitative PET images. Therefore, in order to obtain quantitative PET images, attenuation-corrected maps, also referred to as attenuation maps in the following, are produced, which take into account tissue properties and, if necessary, other factors, such as the MR coil in the imaging area. Attenuation maps can be produced by recording additional image data, such as additional MRT images or CT images and/or from PET scan data. The additional image data is typically recorded together with or shortly before or shortly after the actual primary imaging method.

However, attenuation maps, for their part, often have artifacts that may affect the reliability of the corrected PET image. For example, patient movement, unusual properties of body parts, or other image artifacts may cause erroneous quantification of PET images.

In the prior art, this problem is partly solved by a user, for example an imaging technician or a medical professional, observing and comparing not only images corrected with the aid of attenuation maps but also uncorrected images. However, depending on the user's expertise and attention, some errors, especially minor details, may be overlooked here, which may lead to incorrect diagnosis. Furthermore, the comparison of multiple versions of an image means additional effort for the user. Furthermore, subsequently determined errors may be eliminated by repeating the entire measurement, which results in additional effort and costs. The problems described are particularly acute in the case of PET/MRT, since technicians operating PET/MRT systems, especially nuclear medicine personnel, often do not have particularly outstanding expertise in processing and interpreting the MRT imaging itself. Therefore, artefacts cannot be detected particularly often by the attenuation maps produced by means of MRT.

Contents of the invention

It is therefore an object of the present invention to provide a method for creating quantitative PET images in which the occurrence of artifacts produced by attenuation maps in the PET images can be reduced.

Said object is achieved by a method according to an embodiment of the invention, a positron emission tomography system according to an embodiment of the invention, a computer program according to an embodiment of the invention and a non-volatile computer according to an embodiment of the invention. Read the media to achieve. Further advantages and features emerge from the description and drawings.

According to a first aspect of the invention, a method is proposed for creating quantitative positron emission tomography (PET) images of a subject (Subjekt) or object (Objekt), in particular in PET/MR imaging, PET/CT imaging and / or within the scope of PET imaging. The method includes the following steps:

(a) recording the correction information data by means of at least one imaging device, in particular a magnetic resonance scanner, a computed tomography scanner and/or a positron emission tomography scanner;

(b) sending the correction information data or a first part of the correction information data to the first evaluation unit, by which the first evaluation unit produces an attenuation map based on the correction information data and forwards the attenuation map as inspection data to the second evaluation unit, and/or

forwarding the correction information data or a second part of the correction information data as inspection data to the second evaluation unit;

(c) evaluating the inspection data based on a plurality of reference data by a second evaluation unit and detecting possible artifacts in the inspection data;

(d) In the event that at least one artifact is detected, corrective measures are initialized by the second evaluation unit, and/or

by the second evaluation unit proposing corrective measures to the user via the output device and providing the user with the possibility of initializing corrective measures and/or corrective measures modified by the user,

wherein the correction measure is designed to attenuate and/or eliminate at least one artifact;

(e) Produce a corrected attenuation map based on the correction measures,

wherein the corrected attenuation map corresponds to the uncorrected attenuation map when no artifact is detected; and

(f) Optionally, receiving positron emission tomography measurement data and applying the corrected attenuation map to the positron emission tomography measurement data to create a quantitative positron emission tomography image.

The subject may in particular be a subject or a patient, and the object may be, for example, a body part of the patient.

The first evaluation unit and/or the second evaluation unit can be part of the control unit. The first evaluation unit and/or the second evaluation unit can be arranged on the same computing unit and/or be part of the same evaluation program. In particular, the first evaluation unit and the second evaluation unit may be architecturally combined and/or identical.

Solutions according to the prior art must rely on the evaluation and/or estimation of the user, for example a technician or a doctor, in order to verify the reliability of the PET image, and thus must generally endure increased time expenditure and or an increased probability of error, whereas the method provides a reliable and often faster alternative. Precisely in the case of low user experience, the probability of overlooking image artifacts increases, which in turn may lead to defective and/or erroneous diagnoses. Furthermore, with the aid of the method according to the invention, corrections can possibly be made already during the examination, ie in particular while the patient is still in the imaging scanner. Advantageously, therefore, in the case of an erroneous attenuation map, it may be possible to prevent a repetition of the entire examination by corrective measures or alternatively an evaluation without an attenuation map (ie, a non-quantitative evaluation). The production of attenuation maps or corrected attenuation maps enables a quantitative evaluation of the PET data.

In principle, the method can be applied to different measurement methods including PET imaging. Particularly preferably, the method can be used in the context of PET/MR imaging, PET/CT imaging and/or PET imaging. The correction information data may preferably be recorded by at least one imaging device also used in the actual imaging method. The correction information data can be recorded, for example, by means of a magnetic resonance scanner, a computed tomography scanner and/or a positron emission tomography scanner. It is conceivable that the correction information data are recorded by means of a combination device such as PET/MR, PET/CT, so that the correction information data can also be recorded in multiple modalities and combined therewith. The correction information data can be recorded, for example, by a PET scanner by means of a rotating PET radiation source.

Preferably, the correction information data may be recorded at the beginning of the examination of the subject or object. The correction information data can be recorded as a first scan and/or a first data record directly after recording a positioning image for positioning the area and/or object to be inspected. The correction information data may be, for example, MR data, CT data, PET data, PET/MR data or PET/CT data. Advantageously, therefore, the evaluation or evaluation of the correction information data can still be carried out during the examination, wherein in particular corrections, such as remeasurements of the correction information data, can also still be possible during the examination.

The correction information data are particularly suitable and/or oriented such that an attenuation map can be determined from the correction information data. Within the scope of the present invention, the attenuation map can also be referred to as an attenuation correction map. The term "attenuation" here refers in particular to the absorption of a measurement signal caused by objects, such as tissues and/or other body parts, implants and/or objects arranged between the radiation source and the detector/detectors. Due to individual differences between different subjects, in particular different patients, the absorption of PET signals in different subjects can be different. For example, tissue and/or tissue thickness, such as the thickness of adipose tissue, may vary in different patients. Implants, such as cardiac pacemakers, can also have an effect on the absorption and thus attenuation of the signal and/or cause artifacts.

The correction information data can be sent to the first evaluation unit after recording. Preferably, the correction information data can be forwarded to the evaluation unit directly after recording, ie in particular within a time frame of less than one minute. The first evaluation unit is in particular configured to produce an attenuation map based on the correction information data. Additionally, the first evaluation unit is configured to contain further information, such as PET measurement data, for calculating the attenuation map. The attenuation map can then be sent by the first evaluation unit to the second evaluation unit. Alternatively or additionally, the correction information data can be sent directly, in particular directly, from the imaging device to the second evaluation unit. The correction information data can be used to derive additional and/or more detailed information about possible artifacts. If the correction information data are forwarded directly to the second evaluation unit, it can be provided that the data are checked before the attenuation map is calculated. Advantageously, the calculation time for calculating the attenuation map can therefore be saved if necessary. In particular it can be provided that the attenuation map is only calculated when it has been determined that the correction information data does not have artifacts and/or does not have undesired artifacts. Within the scope of the present invention, the data sent to the second evaluation unit are referred to as inspection data. In other words, the inspection data may include attenuation maps, correction information data, and/or not only attenuation maps but also correction information data. If both the correction information data and the attenuation map are transmitted to the second evaluation unit, the second evaluation unit can be configured to determine the presence of possible artifacts from the combined examination data, ie the correction information data and the attenuation map. Advantageously, artifacts can thus be found particularly reliably. Additionally or alternatively, the attenuation map can be based on a first part of the correction information data, wherein the correction information data forwarded as examination data is based on the entire correction information data or on a second part of the correction information data. In particular, it can be provided that a first part of the correction information data is used to produce the attenuation map, while a second part of the correction information data or the entire correction information data is forwarded directly to the second evaluation unit, so that the second evaluation unit is based on the attenuation map and on the correction information data. The second part or the evaluation is based on the entire correction information data. The first part of the correction information data may include, for example, MR data used to create the attenuation map. The second part of the correction information data may comprise, for example, PET data, in particular PET measurement data. The PET measurement data may be attenuation corrected PET measurement data along with the attenuation map. According to one embodiment, the inspection data may include attenuation corrected PET measurement data and PET measurement data (or uncorrected PET measurement data).

The second evaluation unit may be configured to evaluate the examination data based on a plurality of reference data and to detect possible artifacts in the examination data. The second evaluation unit may be configured to compare the PET measurement data and the attenuation-corrected PET measurement data when evaluating the examination data. Artifacts may be caused, for example, by anomalies in the subject or object and/or by measurement errors. Artifacts may be caused, for example, by missing body parts/limbs, by implants or by objects present in the measurement area. Measurement errors can be caused, for example, by incorrect measurement parameters and/or by (external) interfering signals, such as HF radiation. According to a variant, the second evaluation unit can be configured to use camera data of the subject and/or object provided by the camera for evaluation. The camera data may include in particular information about the movement of the subject and/or the object and/or information about physiological data of the subject and/or the object. The camera can, for example, photograph the subject/object during the examination and/or record images at regular intervals. Physiological data may include, for example, externally identifiable indications of abnormalities of the subject and/or object that cause artifacts or that may cause artifacts.

The second evaluation unit can be configured to initiate corrective measures if at least one artifact is detected. Corrective measures can be initiated, for example, by sending instructions to the first evaluation unit that adapts the evaluation parameters. Alternatively or additionally, a repeated recording of the correction information data or a part of the correction information data can be initiated, wherein optionally, for example, measurement parameters can be adapted. Alternatively or additionally, the second evaluation unit may be configured to propose corrective measures to the user via the output device. It can also be provided that the user is provided with the possibility of initiating corrective measures via the input device and/or of corrective measures being modified by the user. For example, corrective measures can be suggested to the user, the user having the possibility to adapt parameters, such as measurement parameters of repeated measurements, and/or to select alternative corrective measures. Modified corrective measures may also include no corrective measures being provided at all. In other words, it can optionally be provided that the user determines the uncorrected retention of the attenuation map. This can be advantageous, for example, if the user is convinced that the decay map opposite to the recommendation is good enough. It can be provided that an error message is output to the user when at least one artifact is detected. Suggesting corrective measures may advantageously allow the user to be directed to anomalies or artifacts and/or to identify said anomalies or artifacts in the first place. Thus, in addition to the user's expertise and/or attention, a second control mechanism, in particular an independent second control mechanism, can be implemented.

Based on the correction measures, a corrected attenuation map can be produced. The corrective measures may be corrective measures changed by the user. The corrected attenuation map can be produced by the first evaluation unit and/or by the second evaluation unit. Depending on the selected correction measure, a corrected attenuation map can be produced with the help of further measurements by at least one imaging device. If no artifacts are detected, it can advantageously be proposed that absolutely no correction should be applied. Correspondingly, in this case, the corrected attenuation map may correspond to an uncorrected attenuation map.

Positron emission tomography (PET) measurement data may be recorded by an imaging device that also records correction information data. It can particularly preferably be provided that the PET measurement data are recorded in the same examination channel in which the correction information data are also recorded. Thus, several aspects, namely the actual inspection measurements, the recording of the correction information data together with the creation of the attenuation map, and the possible correction of the attenuation map or the correction information data can advantageously be carried out in one inspection channel. Optionally, the PET measurement data may be ToF-PET measurement data (where ToF represents Time of Flight). The signal-to-noise ratio and/or the positional resolution of the measurements can advantageously be improved by means of ToF-PET measurements.

According to a preferred embodiment, the reference data includes in particular a reference attenuation map and/or reference correction information data. The parameter data may in particular come from the same subject or object as the examined object or subject or from the same type of subject and/or object as the examined object or subject. The reference attenuation map or the reference correction information data may in particular come from the object being examined, for example the body region being examined, and/or from an object corresponding to the object being examined, ie in particular a body region of the same type. However, the same type of body region may in particular be the same and/or corresponding body region of another subject or another patient.

According to one embodiment, the first evaluation unit and/or the second evaluation unit may (respectively) comprise a trained algorithm. The trained algorithm can especially be based on artificial intelligence. The trained algorithm of the first evaluation unit and/or the second evaluation unit may be based on deep learning and/or comprise a neural network (NN), for example. A NN may, for example, include a number of 4 to 20 layers, at least some of which may be convolutional layers (English: convolutional layers), and optionally some of the layers may be fully networked layers. The architecture of the NN can be compared, for example, to object recognition NNs such as Joseph Redmon's YOLO.

According to one embodiment, the second evaluation unit is trained by means of input training data and by means of output training data, wherein the input training data includes a plurality of reference data, in particular reference attenuation maps and/or reference correction information data, wherein the output training data includes the input Predetermined classification and/or predetermined evaluation of artifacts in the training data. In particular, the second evaluation unit can be an algorithm trained with the aid of input training data and with the aid of output training data. Multiple reference data can be determined automatically or by the user. The output training data may comprise a classification, in particular a classification determined manually and/or automatically by the user, according to which the reference data is divided into data containing artifacts and data free of artifacts, and/or according to which a classification is determined. The location of the artifact in the reference data. Additionally or alternatively, the output training data may include an evaluation of artifacts in the reference data. The evaluation of artifacts can include, for example, the assignment of significant regions and/or regions of different significant significance, on the basis of which in particular a significance threshold can be determined. Evaluating and/or classifying may include classifying reference data into quality categories. For example, two quality categories may be provided, namely particularly poor, ie with at least one artifact, and good, ie without artifacts. Alternatively or additionally, further quality categories may be provided, for example from very poor, ie in particular very many and/or very noticeable artifacts, to very good, ie in particular no artifacts and/or no noticeable artifacts. graded evaluation. Multiple quality categories can, for example, be used to classify regions of significance.

According to one embodiment, artifacts include anatomical anomalies, in particular missing limbs and/or implants, artifacts due to subject movement, metal artifacts, Gibbs artifacts, aliasing Artifacts or wrap-around artifacts, artifacts due to HF noise and/or artifacts due to chemical offsets. Gibbs artifact is caused especially by the Gibbs phenomenon (Gibbs- ) caused by "Gibbs Ringing", in which additional contours are particularly likely to be produced in the image data. Implants may, for example, cause distortion. Aliasing artifacts may arise, for example, from too small a sampling of the MR measurement, ie in particular too few data points. If the set field of view range of the MR measurement is too small relative to the size of the object being observed, roll-up artifacts can occur, for example, as a sub-form of aliasing artifacts. Here, several parts of the object may appear folded in the field of view. If the different resonance frequencies of water and fat are not taken into account in the MR measurement, artifacts caused by chemical offsets may occur, for example. This may result, for example, in light or dark lines or contours, which may make interpretation difficult and/or erroneous. Additionally or alternatively, artifacts may be caused by a low signal-to-noise ratio, for example due to inappropriate measurement instrument placement and/or coil placement, and/or artifacts may be due in particular to patient Ghosting artifacts caused by insufficient breath holding.

According to one embodiment, the evaluation of the examination data is performed according to a predetermined classification of artifacts based on reference data, in particular reference attenuation maps and/or reference correction information data. The second evaluation unit may be configured to differentiate between multiple categories of artifacts and/or anomalies. Different categories of artifacts can be, for example, anatomical anomalies, artifacts caused by movement of the subject, metallic artifacts, Gibbs artifacts, aliasing or curling artifacts, artifacts caused by RF noise and/or artifacts caused by chemical shifts. A plurality of subcategories may each be provided, which subcategories are further classified into different categories, for example according to the severity of the artifact and/or according to subcategories of the artifact. Subcategories of anatomical anomalies may be, for example, implants and/or missing limbs. Additionally, subcategories of implants may be different types of implants, such as pacemakers or dentures. Thus, advantageously, on the basis of the reference data, for example, known potential error sources can be identified more easily and, if necessary, suitable countermeasures can be introduced, for example corrective measures can be automatically carried out or corrective measures can be suggested.

According to one embodiment, the evaluation is carried out based on a reference attenuation map and/or a reference correction information data based on predetermined evaluation measures, wherein the predetermined evaluation measures include a distinction criterion, in particular a significance threshold, according to which the attenuation map is and/or correction information anomalies in the data are classified as artifacts or no artifacts. The significance threshold can be selected and/or predetermined, for example, depending on the required quality of the respective examination. The significance threshold can be adjusted by the user. For example, a user interface may be provided via which the user can set the significance threshold. The second evaluation unit may be configured to accept user input and/or to adapt the evaluation measure corresponding to the user input.

According to one embodiment, the second evaluation unit includes a series of measures associated with predetermined classification and/or evaluation metrics, some of which include in particular:

-Propose corrective measures to users

- Automatically perform corrective actions

- Optionally, output the information to the user via an output device. Corrective measures can be adapted to the respective artifacts, in particular according to the classification and/or evaluation measures of the artifacts as described herein. Proposing corrective measures to the user and/or corrective measures may include, for example, continuing the scan or inspection, repeating the measurement, in particular repeatedly recording the correction information data, adapting the measurement parameters and/or repeating the scan or inspection. For example, it can be provided that the field of view of the scan is increased, in particular automatically increased, and/or that the phase/frequency encoding is adapted, in particular depending on the severity or prominence of the curling artifacts. Gibbs artifacts that occur can be removed and/or eliminated, for example, by using an increased sampling rate. Corrective measures can be adjusted according to user presets. It can be provided that the proposal of corrective measures and/or the initiation of corrective measures can be adapted or adjusted as a function of the user input. The user input may, for example, apply to all parts of all subsequent measurements or to predetermined parts. It can be provided that the user can be enabled to define a solution strategy for potential artifacts and/or problems that are potential or identifiable by the second evaluation unit, in particular by input via the input device. The solution strategy can, for example, be stored by the second evaluation unit and used and/or taken into account for the initiation of corrective measures and/or for the recommendation of corrective measures. Additionally or alternatively, a significance threshold, especially the significance threshold mentioned above, may be used to determine whether to inform the user about the artifact.

According to one embodiment it can be provided that the correction information data are recorded by a magnetic resonance scanner (MR scanner), in particular part of a combined MR/PET system, the correction information being based in particular on MR-Dixon images. MR-Dixon images may be particularly well suited for avoiding and/or attenuating artifacts caused by erroneous calculation and/or classification of tissue images, in particular regarding the difference between fat and water images. For example, classifications can be set based on air, bone, water, fat, and/or other tissues/objects.

According to one embodiment, when corrective measures are proposed to the user, additional information about the at least one detected artifact, in particular additional information including a graphical representation of the artifact, is output to the user via the output device. Markers that can have artifacts in the image. It may be provided that the marking and/or additional information may be configured by the user. Additionally or alternatively it can be provided that a close-up image with a reduced field of view around the artifact is provided. It can be proposed that artifacts are highlighted by segmentation in their graphical representation. Advantageously, the additional information may provide the user with further support for evaluating and/or identifying artifacts.

According to one embodiment, corrective measures include repeated recording of the correction information data and/or adaptation of recording parameters, in particular recording parameters for the recording of the correction information data to be repeated. Corrective measures can be automatically carried out and/or suggested to the user for manipulation and/or adaptation. The adaptation of the parameters may be based on the classification of artifacts, in particular as described herein. For example, the TE time or another parameter of the MR sequence used in particular to record the correction information data can be changed and in particular can be reduced if a missing signal is identified during the evaluation.

Another aspect of the invention is a positron emission tomography system comprising an imaging device, an output device, an input device and a control unit, wherein the control unit includes a first evaluation unit and a second evaluation unit, and wherein the control unit is configured to operate the imaging Device, wherein the control unit is configured in particular to cause method steps according to the method as described herein, wherein the imaging device is configured to record correction information data, wherein the imaging device is in particular a magnetic resonance scanner, a computed tomography scanner and/or Positron emission tomography scanner, wherein the imaging device is configured to send correction information data to a first evaluation unit, wherein the first evaluation unit is configured to produce an attenuation map based on the correction information data and forward the attenuation map as examination data to a second an evaluation unit, and/or wherein the imaging device is configured to forward the correction information data as examination data to a second evaluation unit, wherein the second evaluation unit is configured to evaluate the examination data based on a plurality of reference data and to detect possible errors in the examination data Artifacts, wherein the second evaluation unit is configured to initiate corrective measures if at least one artifact is detected and/or to propose corrective measures to the user via the output device, wherein the input device is configured to enable the user to implement Initializing a corrective measure and/or a corrective measure modified by the user, wherein the control unit is configured to process the corrective measure input by the user, wherein the corrective measure is designed to attenuate and/or eliminate at least one artifact, wherein the control unit is configured to process the corrective measure based on The correction measure produces corrected correction information data and/or a corrected attenuation map, wherein when no artifact is detected, the corrected correction information data and/or the corrected attenuation map correspond to uncorrected correction information data or An uncorrected attenuation map, wherein the control unit is configured to receive the positron emission tomography measurement data and apply the corrected attenuation map to the positron emission tomography measurement data to create a quantitative positron emission tomography image. All advantages and features of this method can be transferred analogously to positron emission tomography systems and vice versa. The control unit can be, for example, a computer or a part of a computer. The computer can be, for example, a PC, a console of an examination device, a mobile device such as a laptop, a tablet or a smartphone.

Another aspect of the invention is a computer program comprising instructions which, when executed on a control unit of a positron emission tomography system, cause the positron emission tomography system to perform a method as described herein Method steps. All advantages and features of the method and the positron emission tomography system can be transferred analogously to computer programs and vice versa.

Another aspect of the invention is a non-volatile computer-readable medium having stored thereon a computer program, the computer program including instructions that, when used in a control unit of a positron emission tomography system When executed on the computer program, the instructions cause the positron emission tomography system to perform method steps according to a method as described herein. All advantages and features of the methods, positron emission tomography systems and computer programs can be similarly transferred to computer readable media and vice versa. The non-volatile computer readable medium can be any digital storage medium such as a hard drive, server, cloud, computer, optical and/or magnetic storage media, CD-ROM, SSD, SD card, DVD or Blu-ray Disc and/or USB stick .

Unless explicitly stated otherwise, all embodiments described herein may be combined with each other.

Description of the drawings

Embodiments are described below with reference to the drawings.

Figure 1 shows a flowchart describing a method for creating quantitative positron emission tomography images of a subject or object according to the invention,

Figure 2 shows a schematic diagram of a positron emission tomography system according to the invention.

Detailed ways

Figure 1 shows a flowchart describing a method for creating quantitative positron emission tomography images (PET images) of a subject or object according to the present invention. The method may preferably be applied during PET/MR imaging, PET/CT imaging and/or PET imaging. In a first step 11 , correction information data is recorded by means of at least one imaging device 2 . The imaging device 2 can be, for example, a magnetic resonance scanner, a computed tomography scanner and/or a positron emission tomography scanner. In the following step 12, the correction information data or a first part of the correction information data can be sent to the first evaluation unit. In step 13, the first evaluation unit creates an attenuation map based on the correction information data. The attenuation map is then further processed as inspection data. Alternatively or additionally, the correction information data themselves or a second part of the correction information data can also be processed as test data.

In a next step 14, the inspection data are forwarded to the second evaluation unit. In step 15 , the examination data are evaluated by the second evaluation unit on the basis of a plurality of reference data and possible artifacts in the examination data are detected. The reference data may preferably include a reference attenuation map and/or reference correction information data. Artifacts may include, for example, missing limbs and/or implants, artifacts caused by movement of the subject, metal artifacts, Gibbs artifacts, aliasing artifacts or wrap artifacts. -Around-Artefakte), artifacts caused by HF noise and/or artifacts caused by chemical shifts. The evaluation of the examination data can be performed according to a predetermined classification of artifacts and/or according to a predetermined evaluation measure. Preferably, the evaluation measure may comprise a discrimination criterion, in particular a significance threshold, according to which anomalies in the attenuation map and/or the correction information data are classified as artifacts or not artifacts.

The first evaluation unit and/or the second evaluation unit may be part of the control unit 5 and/or comprise a trained algorithm. In particular, the second evaluation unit can be trained by means of input training data and by means of output training data. The input training data may include a plurality of reference data, in particular reference attenuation maps and/or reference correction information data, wherein the output training data may include a predetermined classification and/or a predetermined evaluation of artifacts in the input training data. .

If no artifacts are found, proceed directly to step 18. For the case where at least one artifact is detected, corrective measures are initiated in step 16 by the second evaluation unit. Alternatively or additionally, in step 17 the second evaluation unit proposes corrective measures to the user via the output device 3 and provides the user with the possibility of initializing corrective measures and/or corrective measures modified by the user. Optionally, it can be provided that, when proposing corrective measures to the user, additional information about at least one detected artifact, in particular additional information including a graphical representation of the artifact, is output to the user via the output device 3 . Graphical representations of artifacts may include, for example, segmentation of artifacts. In this case, the correction measure is designed to attenuate and/or eliminate at least one artifact. The second evaluation unit may in particular comprise a series of measures associated with predetermined classification and/or evaluation measures. Corrective measures may include in particular repeated recording of the correction information data and/or adaptation of recording parameters for the recording of the correction information data to be repeated.

In step 18, a corrected attenuation map is produced based on the corrective measures. If no artifacts are detected, the corrected attenuation map may correspond to the uncorrected attenuation map. Finally, in step 19, the positron emission tomography measurement data are received and the corrected attenuation map is applied to the positron emission tomography measurement data in order to create a quantitative positron emission tomography image.

Figure 2 shows a schematic diagram of a positron emission tomography system 1 according to the invention. The imaging device 2 of the system 1 is configured for recording correction information data. The imaging device 2 may in particular be a magnetic resonance scanner, a computed tomography scanner and/or a positron emission tomography scanner. The imaging device 2 can be configured to perform different imaging methods, in particular PET imaging and another imaging, such as MR imaging. The imaging device 2 is further configured to send correction information data to the first evaluation unit. In this embodiment, the first evaluation unit together with the second evaluation unit is part of a control unit 5 which is configured for operating and/or controlling the imaging device 2 . The control unit 5 is configured in particular to carry out the method steps as described in FIG. 1 or to cause the execution of said method steps. The method steps may preferably be stored as a computer program on a computer-readable medium, which may in particular be part of the control unit or be connected to the control unit.

The first evaluation unit is configured to produce an attenuation map based on the correction information data obtained by the imaging device 5 and to forward the attenuation map as inspection data to the second evaluation unit. Alternatively or additionally, the imaging device is configured to forward the correction information data directly to the second evaluation unit as examination data.

The second evaluation unit is configured as part of the control unit 5 for evaluating the examination data on the basis of a plurality of reference data and for detecting possible artifacts in the examination data. If the second evaluation unit detects at least one artifact, it is configured to initiate corrective measures for eliminating and/or attenuating the artifacts and/or to propose corrective measures to the user via the output device 3 . The user can initiate corrective measures and/or corrective measures modified by the user via the input device 4 . The control unit 5 is here configured to process corrective measures input by the user. Furthermore, the control unit 5 is configured to produce corrected correction information data and/or a corrected attenuation map based on corrective measures, for example by means of a first evaluation unit, and to receive the PET measurement data and apply the corrected attenuation map to the PET measurement. data in order to create quantitative PET images.

Claims (13)

1. A method for creating quantitative positron emission tomography images, PET images of a subject or object, in particular in the scope of PET/MR imaging, PET/CT imaging and/or PET imaging, the method comprising the steps of:

(a) Recording correction information data by means of at least one imaging device (2), in particular a magnetic resonance scanner, a computer tomography scanner and/or a positron emission tomography scanner;

(b) Transmitting the correction information data or a first part of the correction information data to a first evaluation unit, producing an attenuation map based on the correction information data by the first evaluation unit and forwarding the attenuation map as inspection data to a second evaluation unit, and/or

Forwarding the correction information data or a second part of the correction information data as inspection data to the second evaluation unit;

(c) Evaluating, by the second evaluation unit, the inspection data based on a plurality of reference data and detecting possible artifacts in the inspection data;

(d) In the event of detection of at least one artifact, initiating corrective measures by means of the second evaluation unit, and/or

By the second evaluation unit making corrective measure suggestions to a user via an output device (3) and providing the user with the possibility of initializing the corrective measures and/or corrective measures modified by the user,

wherein the corrective measure is designed to attenuate and/or eliminate the at least one artifact;

(e) A corrected attenuation map is made based on the corrective measure,

wherein when no artifact is detected, the corrected attenuation map corresponds to an uncorrected attenuation map; and

(f) Optionally, positron emission tomography measurement data is received and the corrected attenuation map is applied to the positron emission tomography measurement data to create a quantitative positron emission tomography image.

2. The method according to claim 1,

wherein the reference data comprises a reference attenuation map and/or reference correction information data.

3. The method according to claim 1 or 2,

wherein the first evaluation unit and/or the second evaluation unit comprises a trained algorithm.

4. The method according to any of the preceding claims,

wherein the evaluation is performed according to the reference attenuation map and/or the reference correction information data according to a predetermined classification of artifacts.

5. The method according to any of the preceding claims,

wherein the evaluation is performed according to a predetermined evaluation metric in accordance with a reference attenuation map and/or reference correction information data,

wherein the predetermined evaluation measure comprises a differentiation criterion, in particular a significance threshold, according to which anomalies in the attenuation map and/or the correction information data are classified as artifacts or are free of artifacts.

6. The method according to claim 4 or 5,

wherein the second evaluation unit comprises a series of measures, which measures are associated with the predetermined classification and/or the evaluation measure,

wherein the series of measures comprises in particular:

-suggesting said corrective measures to said user-automatically performing corrective measures

-optionally, outputting information to the user via the output device (3).

7. The method according to any of the preceding claims,

wherein the artifacts comprise anatomical abnormalities, especially absent limbs and/or implants, artifacts caused by movement of the subject, metal artifacts, gibbs artifacts, aliasing or crimping artifacts, artifacts caused by HF noise and/or artifacts caused by chemical shifts.

8. The method according to any of the preceding claims,

wherein when a suggestion of a corrective measure is made to a user, additional information about the at least one detected artifact, in particular additional information comprising a graphical representation of the artifact, is output to the user via the output device (3).

9. The method according to any of the preceding claims,

wherein the corrective measure comprises repeated recording of the correction information data and/or adapting recording parameters for the recording of the correction information data to be repeated.

10. The method according to any of the preceding claims,

wherein the second evaluation unit is trained by means of the input training data and by means of the output training data,

wherein the input training data comprises a plurality of reference data, in particular reference attenuation maps and/or reference correction information data,

wherein the output training data comprises a predetermined classification and/or a predetermined evaluation of artifacts in the input training data.

11. A positron emission tomography system (1), the positron emission tomography system (1) comprising an imaging device (2), an output device (3), an input device (4) and a control unit (5),

wherein the control unit (5) comprises a first evaluation unit and a second evaluation unit, and wherein the control unit (5) is configured for controlling the imaging device (2),

wherein the control unit (5) is especially configured for causing the method steps of the method according to any one of the preceding claims,

wherein the imaging device (2) is configured to record the correction information data,

wherein the imaging device (2) is in particular a magnetic resonance scanner, a computer tomography scanner and/or a positron emission tomography scanner,

wherein the imaging device (2) is configured to send the correction information data to the first evaluation unit,

wherein the first evaluation unit is configured to produce an attenuation map based on the correction information data and to forward the attenuation map as inspection data to the second evaluation unit, and/or

Wherein the imaging device (2) is configured to forward the correction information data as inspection data to the second evaluation unit,

wherein the second evaluation unit is configured for evaluating the examination data based on a plurality of reference data and for detecting possible artifacts in the examination data,

wherein the second evaluation unit is configured to initiate the corrective measure in the event of detection of at least one artifact, and/or

Suggesting said corrective measures to the user via said output device (3),

wherein the input device (4) is configured for enabling a user to initiate the corrective action and/or a corrective action modified by the user, wherein the control unit (5) is configured for processing the corrective action entered by the user,

wherein the corrective measure is designed to attenuate and/or eliminate the at least one artifact,

wherein the control unit (5) is configured for producing corrected correction information data and/or corrected attenuation maps based on the correction measures,

wherein when no artifact is detected, the corrected correction information data and/or the corrected attenuation map corresponds to uncorrected correction information data or uncorrected attenuation map,

wherein the control unit (5) is configured for receiving the positron emission tomography measurement data and applying the corrected attenuation map to the positron emission tomography measurement data in order to create a quantitative positron emission tomography image.

12. A computer program comprising instructions which, when executed on a control unit (5) of a positron emission tomography system (1), cause the positron emission tomography system (1) to perform the method steps of any one of claims 1 to 10.

13. A non-transitory computer readable medium on which a computer program is stored, the computer program comprising instructions which, when executed on a control unit (5) of a positron emission tomography system (1), cause the positron emission tomography system (1) to perform the method steps of any one of claims 1 to 10.